TEXT ANALYTICS SOLUTIONS
TAS Enterprise Search Engine
TAS Enterprise Search developed by Precognox is a state-of-the-art search engine that offers theme-independent, intelligent, semantic search. The solution ensures the user to search multiple large databases at once as the search engine can be connected to a wide range of data sources of the company. By advanced search functions and operators specially compiled queries can be run to solve complex search tasks. Bulk search option simplifies time-consuming searches which were previously done individually. The theme-independent search capabilities are supported by the customizable built-in synonym dictionary (thesaurus) to find relevant results. Leading entity recognition and name matching solutions are also integrated.
TAS Data Collector
TAS Data Collector allows to collect all unstructured and structured data from a certain or more domain on the Internet. The interface provides the ability to monitor data flow continuously so that daily data trends can be monitored. The collected data can be used in raw form, or converted and utilized by the further solutions of the TAS Text Analytics System.
TAS Search Log Analyzer
TAS Search Log Analyzer is a supplementary service for tracking and analyzing queries made in the search engine. This analytical tool provides insight into which search terms are used and how they are used in TAS Enterprise Search Engine interface. The data obtained can also be useful to improve the performance of the enterprise search engine.
TAS Thesaurus Manager
TAS Thesaurus Manager is a thesaurus building module that enables to define and verify conceptual relations (synonym, typo, correct form) between the word pairs. The defined relations make it easier to find relevant results in TAS Enterprise Search Engine. Thus the solution increases the efficiency of the search process and reduces the number of queries.
TAS Tagger retrieves and determines key phrases and topics from text contents. The identification of these expressions and named entities (person names, locations, organizations) are implemented by computational linguistic and machine learning methods and tools. We apply some or all of these methods and tools. It depends on the requirements of the particular customer and the given task. Text contents can be automatically tagged (labelled) by the extracted and determined key phrases.
Research and development of AI (Artificial Intelligence) is one of the most important and fastest growing fields. AI opens up new perspectives on text analytics, thus application of Artificial intelligence is essential and plays a main role in our solutions, such as data and text mining, text analytics or computational language processing.
Optical character recognition
Optical character recognition technology (OCR) is applied to extract text contents from image files and digitized image documents. The extracted text contents are already suitable for use in our text analytics solutions. The integrated OCR technology is provided by ABBYY FineReader. We have dedicated expert for processes that need OCR technology.
Our company is an ABBYY reseller partner.
Natural language processing
Natural language processing is a common subset of artificial intelligence and linguistics, where natural languages are processed by computational methods. One of the main pillars of our services is the so called NLP (Natural language processing), which serves as a basis for many applications, such as content analysis, text tagging, auto-translation or sentiment analysis. In natural language processing our partner is Basistech, the developer of the Rosette text analytics system. Rosette offers integrated solutions such as entity extraction, name matching, language identification, or sentiment analysis.
Machine learning is the subfield of artificial intelligence. We use machine learning technologies to improve the efficiency of our solutions: we teach our algorithms through learning data to produce an output that contains predictions based on the data. We are also able to create models for machine learning solutions and to prepare the necessary datasets accordingly.